The ghostly gel has been caught. The chemistry is no longer invisible.
For years, the concrete industry has been locking carbon dioxide into its mixes without truly understanding what transpired inside the hardening paste — an act of intuition rather than knowledge. Now, researchers at MIT have used laser light to illuminate a three-stage chemical drama unfolding in the first hours of cement's life, revealing how CO2 temporarily rewrites the material's inner architecture before vanishing, leaving behind a stronger, more uniform structure. The discovery, born from watching carbon dioxide turn to snow in a Cambridge laboratory, transforms an empirical practice into a legible mechanism — and opens the door to deliberate optimization of a process that could meaningfully reduce one of industry's most stubborn sources of emissions.
- Cement producers have been injecting CO2 into concrete for years to sequester carbon, but the chemistry driving the strength gains was effectively invisible — a black box at the heart of a growing industrial practice.
- MIT researchers cracked that black box open using Raman confocal microscopy, watching in real time as CO2 triggered a cascade of reactions — calcium carbonate formation, silica gel dispersal, and finally the creation of evenly distributed binding compounds — across a critical eight-hour window.
- The 13% jump in compressive strength at 24 hours turns out to hinge on a temporary drop in alkalinity: CO2 suppresses the paste's pH just long enough for silica gel to spread uniformly before reacting into the binding compound that gives cement its strength.
- The finding reframes what was thought to drive the effect — calcium carbonate crystals, once suspected as active strength-builders, are now understood to be passive scaffolding, while the fleeting silica gel network does the real structural work.
- The mechanism also defines the limits: too much CO2 overwhelms the system and locks calcium away before the gel can form, meaning optimization — not maximization — is the path forward toward an estimated 40% emissions offset.
On a September afternoon at MIT's Pierce Laboratory, researchers watched liquid CO2 depressurize into solid flakes, then mixed those frozen particles into fresh cement paste. They pressed the mixture into dime-sized disks, sealed them with vegetable oil, and pointed lasers at them — observing, for the first time, the chemical reactions unfolding inside hardening cement in real time. Over 24 continuous hours, they mapped a three-stage process that explains why CO2-injected concrete gains strength faster than conventional mixes.
The work, led by associate professor Admir Masic and graduate student Marcin Hajduczek and published in the Journal of the American Ceramic Society, addressed a puzzle the industry had long set aside. Companies have injected CO2 into concrete for years as a carbon-sequestration strategy, but the internal chemistry moved too quickly and left too few traces for older techniques to capture. Raman confocal microscopy changed that: shine a laser on a molecule, and the scattered light reads its chemical identity like a fingerprint, catching even the most transient phases.
The first act begins the moment CO2 enters fresh paste. The gas dissolves into pore liquid and reacts with calcium from dissolving clinker, forming calcium carbonate crystals that lock the calcium away. Starved of calcium, silicates spread throughout the paste as a thin, amorphous gel — a ghostly network threading the entire matrix within the first hour. Around four to five hours in, the CO2 is fully mineralized and normal hydration resumes. Calcium hydroxide precipitates into the pores and immediately encounters the waiting silica gel, reacting to form calcium silicate hydrate — the compound that actually binds cement — distributed evenly across the matrix rather than clustered around clinker particles. By eight hours, the gel has nearly vanished, transformed into additional binder during that critical early window.
The result is a measurably more uniform microstructure. Cement paste with CO2 at one percent by weight achieved 13% higher compressive strength at 24 hours compared to reference mixes. Crucially, the calcium carbonate crystals long suspected of driving the effect turned out to be passive — embedded in the silica gel template rather than building strength themselves.
The practical ceiling is real but meaningful. Too much CO2 defeats the process by locking calcium into carbonate before the gel can form and react. If optimized, the mechanism could theoretically offset up to 40% of cement production's carbon emissions — though that figure excludes the fossil fuels burned in production, and the achievable real-world fraction is likely smaller. Next steps include directly measuring the silica gel's mechanical properties and refining dosage. The ghostly chemistry that unfolds in cement's first hours is no longer invisible, and Masic sees considerable room to push.
On a September afternoon at MIT's Pierce Laboratory, researchers watched carbon dioxide turn to snow. They depressurized a tank of liquid CO2, and solid flakes tumbled out—the kind of thing you might see in a high school chemistry demo, except what came next would reshape how scientists understand cement. The team mixed those frozen particles into fresh cement paste, pressed the mixture into disks the size of a dime, sealed each one with vegetable oil, and pointed lasers at them. For the first time, they could see the chemical reactions happening in real time as the paste hardened. What they discovered over 24 hours of continuous observation was a three-stage process that explains why CO2-injected concrete gets stronger faster than the conventional kind.
The work, published in the Journal of the American Ceramic Society and led by MIT associate professor Admir Masic and graduate student Marcin Hajduczek, answers a question that had puzzled the industry for years. Companies have been injecting CO2 into concrete mixes for a while now—it's a way to lock carbon away and keep it out of the atmosphere—but nobody had actually seen what was happening inside the material as it set. The chemical reactions moved too fast and left too few traces for older measurement techniques to catch. Raman confocal microscopy changed that. Shine a laser on a molecule, and the scattered light reveals its chemical identity like a fingerprint. Even the most fleeting, ghostly phases leave a readable mark.
The first act begins the moment CO2 hits fresh cement paste. The gas dissolves into the liquid filling the pores and reacts with calcium released by the dissolving clinker—the mineral powder that forms the base of all cement. This reaction produces calcium carbonate, which precipitates out of solution. Normally, that calcium would stay nearby and feed the paste's hydration process, the slow chemical transformation that makes cement hard. But with CO2 present, the calcium gets locked away into carbonate crystals instead. Starved of calcium, the silicates released by the clinker dissolve and spread throughout the paste, forming a thin, amorphous gel network that threads through the entire matrix. This happens in the first hour, and it's the setup for everything that follows.
Around four to five hours after mixing, the injected CO2 is fully converted to mineral form, and normal hydration resumes. Calcium hydroxide begins to precipitate into the pores, and when it does, it encounters the silica gel network that has been waiting there. The two react immediately, producing calcium silicate hydrate—the compound that actually gives cement its strength and binding power. But here's what makes this different from conventional cement: the new binder doesn't cluster around the original clinker particles. Instead, it forms wherever the silica gel had spread, distributed evenly throughout the entire matrix. The CO2 had temporarily lowered the paste's alkalinity, and that lower pH was the only thing keeping the silica gel intact. As hydration reasserts itself and raises the pH back up, the gel reacts away almost completely within eight hours. The previously ghostly network transforms into additional binding compound in a critical early window.
By the time the paste has fully set, the microstructure is measurably different from conventional cement. Because the new binder was distributed more evenly rather than clustered, the resulting material is stronger and more uniform at an early age. In the study, cement paste mixed with CO2 at one percent by weight achieved, on average, 13 percent higher compressive strength at 24 hours compared with reference mixes. That's a significant jump in a short time. The calcium carbonate crystals that researchers had previously suspected of actively building strength turned out to be passive passengers embedded in the silica gel template, not the drivers of the reaction.
Masic notes that the team has been using Raman spectroscopy to understand some of history's most interesting materials—ancient Roman concrete, the Dead Sea Scrolls—but pointing a laser at CO2-injected cement paste as it hardens revealed something entirely new. "We've been injecting CO2 into cement products for years without fully understanding what it was doing inside," he says. "Now that we can see it and understand the underlying mechanism that leads to improved performance, we can start to control it. And there's a lot of room to push."
The practical implications are significant but bounded. Too much CO2 floods the system and locks calcium into carbonate before the gel can form and react, which defeats the purpose. If the process works as theoretically optimized, it could offset up to 40 percent of the carbon emissions from cement production—though that calculation excludes the fossil fuels burned to create the heat needed in the first place. In practice, the achievable offset is likely to be only a fraction of that theoretical maximum, though still potentially meaningful. The next steps involve directly measuring the mechanical properties of the silica gel itself and refining the dosage to maximize the benefit without overshooting. But the ghostly gel has been caught. The chemistry that unfolds in those first eight hours is no longer invisible.
Notable Quotes
We've been injecting CO2 into cement products for years without fully understanding what it was doing inside. Now that we can see it and understand the underlying mechanism, we can start to control it.— Admir Masic, MIT associate professor
The Hearth Conversation Another angle on the story
Why does it matter that we can now see this happening? Couldn't researchers already predict the outcome?
They could predict some of it from theory, but seeing it directly changes everything. It's like the difference between reading a recipe and watching someone actually cook. The silica gel was so fleeting that conventional techniques couldn't catch it—it appeared and disappeared too fast. Raman spectroscopy let them watch the whole sequence unfold.
So the CO2 is doing something unexpected—it's not just filling pores or reacting directly with the cement?
Exactly. The CO2 temporarily starves the paste of calcium by locking it into carbonate crystals. That forces the silicates to spread out and form a gel network throughout the entire matrix instead of clustering locally. It's a detour that leads to a better final structure.
And that gel network—it doesn't stay around?
No, it's consumed within eight hours. As normal hydration resumes and the pH rises, the gel reacts with calcium hydroxide and transforms into the binding compound. But by then it's already done its job: it's distributed the new binder evenly throughout the paste instead of leaving it clustered.
Is 13 percent stronger at 24 hours actually significant in real construction?
It's meaningful. Early strength matters for things like precast concrete products that need to be moved quickly, or for structures that need to bear load soon after pouring. But the bigger significance is that now they understand the mechanism, they can optimize it. They might be able to push the effect further.
What's the carbon story here? Is this actually a way to reduce emissions?
It can be, but with limits. If optimized perfectly, it could offset up to 40 percent of cement production emissions. But that's theoretical. In practice, you have to account for the energy used to capture and inject the CO2 in the first place. The real offset is likely much smaller, though still potentially significant enough to matter at scale.